Cooperative beamforming in multihop amplify-and-forward relay networks

This paper studies cooperative beamforming in multihop amplify-and-forward relay networks. We adopt a layer-by-layer approach, namely per-layer sum-rate maximization, to compute the beamforming weights employed by the relays at each layer. When there is only one receiving node at the adjacent downstream layer, the per-layer sum-rate maximization problem is equivalently converted into a convex quadratically constrained quadratic program. When there are multiple receiving nodes at the adjacent downstream layer, the per-layer sum-rate maximization problem is nonconvex and is equivalently transformed into a difference-of-convex (DC) program. We develop a successive convex approximation algorithm to approximately solve this DC program. Numerical results confirm that the proposed beam-forming design outperforms the existing ones in the literature in terms of achievable data rate, and the proposed scheme is effective for suppressing noise propagation.